Last Friday Kent Erickson, Deepak Kanwar, and David Winter of Zenoss hosted a webinar with the unassuming title Optimize IT Operations for Data Center Efficiency, which for the most part felt like a review session of the major topics discussed on this blog over the past few months. For the first two-thirds of the webcast, Kent et al. discussed the problems caused by the proliferation of infrastructure monitoring tools, the resistance within IT shops in breaking down IT silos, and the inability of many organizations to maintain even 99.0% uptime for its mission-critical applications. If you’re unfamiliar with these issues or just need a refresher, this portion of the webinar will give you a good primer.
But for those of you who have already read the Forrester Survey and know the benefits of converged infrastructure (among other things), jump to 37:45 because that is where, for me anyway, the webinar starts to get exciting. Kent starts this section by asking Dave to describe what you need to run Zenoss Service Dynamics, the company’s flagship product, on-premise. Then Kent asks Dave to contrast that with what you need to run ZaaS (Zenoss as a Service), which was introduced just a week ago. Dave’s team has been building this hosted version of ZSD over the last year to give potential customers what he called “a new consumption model...of the Zenoss we all know and love.”
Between you and me, I’m perplexed by this under-the-radar launch of ZaaS. I’ve always been onboard with Zenoss’ unified monitoring approach because I’ve interviewed too many frustrated IT people over my 15 years of covering enterprise technology. But I had been skeptical about the on-premise version of ZSD being a realistic solution either for mid-market companies with limited resources or for newer companies that may have built most, if not all, of their infrastructure in a private or public cloud. Why would these companies use any sort of on-premise solution, even one as useful as Zenoss?
They buried the lead
Because ZaaS’s introduction got a bit buried in the webinar, let me bring you back to Deepak’s exuberant post ZaaS What I’m Talking About!, where he stresses ZaaS is not simply a “Lite” version of ZSD:
Vendors, in their bid to be early entrants, have brought to market dumbed-down and often highly diluted versions of their flagship products. While quick to deploy, the capabilities of these solutions have left a lot to be desired...
We believe that the robustness of the solution should not be compromised, regardless of the consumption model...and that was the thinking behind offering Zenoss Service Dynamics as a Service. Zenoss as a Service–“ZaaS” as we like to call it–allows customers to take advantage of the advanced capabilities found in our on-premise solution, without owning the process of deployment, support or management. For some, a hosted solution could mean the difference between deploying unified monitoring and sticking with sub-optimal monitoring.
Even the patent-pending Service Impact capabilities are available as a service, meaning you start troubleshooting any infrastructure issue with a rank-ordered triage list. The one thing Kent and the team said isn’t currently available with ZaaS is the “Big Data”-style Analytics module – typically used by larger enterprises for quarter-over-quarter or year-over-year capacity forecasting. But those organizations probably have considerable IT staffing who manage a lot of on-premise tools.
In my follow-up with Deepak post-webinar, he did share one other point you may want to consider when weighing your options. As is the case with most SaaS services, ZaaS does store monitoring data in the cloud. If that’s against your policies, no ZaaS for you!
But those of you who are already relying on cloud-based infrastructures or hybrid infrastructures have another option now. Even if you have just one person handling your IT, you want assurance that everything is online and responsive to your customer or end-user needs just like the big guys. And because your business isn’t “too big to fail,” even the mildest glitch can cripple it.
Which is why ZaaS really is a big deal. It’s akin to what Salesforce.com achieved a decade ago. Before Salesforce.com and other SaaS services came to market, you could either install cumbersome CRM applications from big vendors like PeopleSoft and Siebel (remember them?) or you made do with some jury-rigged mashup of Microsoft Outlook and Excel. Similarly with ZaaS, you no longer have to put your hopes into a mishmash of monitoring tools that lack even common frames of reference.
The Joy of ZaaS
During the webinar, Dave explained that ZaaS is deployed in the cloud and works with your infrastructure through a secure connection to an on-premise virtual “collector” that is installed into your hypervisor:
[This collector does] all the unified monitoring that ZSD does, be it SNMP-based, SSH-based, or specific API-based, [along with] log collection and analysis. You can still run the gamut of ZenPacks and [take advantage of] the extensibility of Zenoss, scalable through multiple collectors, but the back-end administration and troubleshooting is being handled by us, the manufacturer, in a well-designed, formally maintained environment.
What does Dave mean when he says “handled by us?” Here’s a partial list:
- Initial Deployment
- Maintenance (including patches and upgrades)
- Infrastructure support
- Tech support
- Disaster recovery
- ZaaS application health (which includes a 99.9% uptime SLA)
Dave said the motivation behind building ZaaS was this:
Zenoss is taking on the responsibility of providing you, the customer, with a clean stable system you don't have to worry about. We’re relieving this burden so that you can just use Zenoss.
All I could think in that moment was, Yes!
Zenoss’ best quality is its attention to simplicity. It’s about the hardest thing to achieve in any domain, be it writing, music, or software development. The fact that the company has built and continues to develop a unified monitoring solution “from the ground up” that can monitor most anything in a data center is ground-breaking in its own right. But providing a version most any organization can use, will, at the very least, change the way we do business going forward – at least in my modest opinion...